**** BEGIN LOGGING AT Fri Oct 03 03:00:00 2014
Oct 03 04:33:13 <Devastator> morning :)
Oct 03 04:39:27 <zajec> Devastator: hey, you live in Europe? USA?
Oct 03 04:39:28 <zajec> ;)
Oct 03 04:39:37 <Devastator> Brazil haha
Oct 03 04:39:54 <Devastator> but I'm used to Europe timezone, so.. it's morning! hahah
Oct 03 04:40:18 <zajec> :)
Oct 03 04:40:36 <Devastator> are you an operator here or #openwrt?
Oct 03 04:40:47 <Devastator> I think the topic could be changed after BB release
Oct 03 04:41:14 <zajec> i'm not
Oct 03 04:41:21 <zajec> but good point
Oct 03 04:43:19 <Devastator> zajec do you happen to build your own images?
Oct 03 04:53:31 <zajec> Devastator: sure, how else I could commit anything? ;)
Oct 03 04:56:16 <Devastator> zajec heheh true.. I would like to ask you a favour, or at least some guidance: I'm trying very hard to build my own image with a bigger rootfs but kernel md5sum changes everytime and then I can't use official repos because kernel doesn't match, do you have any idea or can try to build one image for me with the same kernel md5sum as the official images?
Oct 03 04:58:16 <zajec> 1) why rootfs doesn't use the whole space for you?
Oct 03 04:58:38 <zajec> 2) does OpenWrt really refuse to use repos because of mismatching md5? there must be a way to change it
Oct 03 04:59:09 <zajec> Devastator: pastebin  your boot log
Oct 03 05:00:19 <Devastator> zajec because cavium octeon only have a generic profile yet
Oct 03 05:00:51 <zajec> paste bin boot log
Oct 03 05:00:55 <zajec> i have no idea of this platform
Oct 03 05:01:00 <zajec> need to see how it works
Oct 03 05:01:40 <Devastator> zajec can it be from the image it is currently running?
Oct 03 05:01:45 <Devastator> it has the mismatch problem :P
Oct 03 05:02:11 <zajec> sure
Oct 03 05:03:27 <Devastator> zajec codepad.org/MefPXpRE
Oct 03 05:03:33 <Devastator> opss
Oct 03 05:03:38 <Devastator> zajec http://codepad.org/MefPXpRE
Oct 03 05:07:32 <zajec> oh, it has a normal hard driver for storage
Oct 03 05:08:00 <zajec> oh, it's a USB disk
Oct 03 05:08:02 <zajec> but still
Oct 03 05:08:43 <zajec> does this USB has one partition with squashfs.img on it?
Oct 03 05:10:23 <Devastator> zajec two, one is fat with kernel and kernel md5, rootfs is in another ext4 partition
Oct 03 05:12:29 <Devastator> when I first flashed this router it had a 148MB FAT partition, a 2GB ext4 and 1.7GB unallocated, then I repartitioned it to 3.7GB ext4
Oct 03 05:15:18 <zajec> ok, unfortuantely i don't know how partitioning works for USB drives :|
Oct 03 05:15:31 <zajec> but you should make OpenWrt use packages repo anyway I believe
Oct 03 05:15:34 <zajec> Google for the error you get
Oct 03 05:18:49 <Devastator> zajec I think normal packages are ok, only kmod-packages or kernel related packages that gives headaches :)
Oct 03 05:19:18 <zajec> ok, that may be expected
Oct 03 05:19:25 <zajec> but you still should be able to force that
Oct 03 05:24:45 <Devastator> zajec if I use --force-depends it install
Oct 03 05:25:09 <Devastator> but well, would be nice to have something close to official ones
Oct 03 05:41:32 <Devastator> zajec thank you for your help
Oct 03 05:42:15 <Devastator> I wonder if there is a .config somewhere that matches the md5sum of official builds, than I can compare
Oct 03 06:23:16 <builder> build #194 of adm8668 is complete: Failure [failed compile_5]  Build details are at http://buildbot.openwrt.org:8010/builders/adm8668/builds/194
Oct 03 09:14:48 <builder> build #786 of brcm63xx is complete: Failure [failed compile_8]  Build details are at http://buildbot.openwrt.org:8010/builders/brcm63xx/builds/786
Oct 03 09:19:25 <builder> build #705 of ppc44x is complete: Failure [failed shell_10]  Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/705
Oct 03 10:01:29 <karlp> if you're building your own images, why do you need to be able to use the standard package repos?
Oct 03 10:29:56 <zajec> Hauke: hi :) you missed some recent bcm53xx fun
Oct 03 10:30:21 <zajec> (i mean changes over last 2 days afair)
Oct 03 10:41:01 <OpenWrt-Commits> rmilecki r42754 trunk/target/linux/bcm53xx/image/Makefile * bcm53xx: build R6250 firmware with "Other Regions" region
Oct 03 10:57:30 <OpenWrt-Commits> cyrus r42755 trunk/package/base-files/files/etc/uci-defaults/12_network-generate-ula * base-files: remove a 'not found' error message during system boot
Oct 03 11:04:45 <OpenWrt-Commits> jogo r42756 trunk/target/linux/brcm63xx/patches-3.14/505-board_spw500v.patch * brcm63xx: fix wrong fixups pointer
Oct 03 11:05:19 <Hauke> zajec: I was busy with non linux/openwrt related stuff
Oct 03 11:05:30 <Hauke> what happened?
Oct 03 11:05:45 <zajec> i've backported upstreamed bcma patches
Oct 03 11:05:51 <zajec> sent some ARM patches for inclusion
Oct 03 11:05:58 <zajec> not sure if they will make 3.18
Oct 03 11:06:06 <Hauke> probably not
Oct 03 11:06:07 <zajec> added GPIO chip support
Oct 03 11:06:12 <zajec> added LEDs for R6250
Oct 03 11:06:18 <zajec> and kind-of buttons support
Oct 03 11:06:18 <Hauke> nice
Oct 03 11:06:25 <zajec> however it's polled... not working that cool
Oct 03 11:06:32 <zajec> i used 200ms polling interval
Oct 03 11:06:51 <Hauke> but using irqs should not be so hard?
Oct 03 11:07:01 <zajec> in theory
Oct 03 11:07:13 <zajec> in practice i couldn't get any button IRQ, no idea why
Oct 03 11:07:26 <zajec> i was getting UART IRQs for every char
Oct 03 11:07:30 <zajec> but not buttons
Oct 03 11:07:33 <zajec> have to work more on that
Oct 03 11:07:41 <Hauke> hmm strange
Oct 03 11:07:57 <Hauke> could it be that gpio uses a different irq?
Oct 03 11:08:24 <Hauke> irqs are not limited any more, the nand core has 8 irqs
Oct 03 11:09:11 <zajec> i got a bit lost in buttons IRQ
Oct 03 11:09:15 <zajec> i was looking at this arch code
Oct 03 11:09:43 <zajec> I think what Broadcom did in arch code is... well, re-implementation of IRQ domain
Oct 03 11:09:52 <zajec> they have many interrupts that don't seem to be coming from the hw
Oct 03 11:12:12 <Hauke> zajec: can you point he to the file
Oct 03 11:12:30 <zajec> sec
Oct 03 11:19:54 <zajec> Hauke: R8000-V1.0.0.46_1.0.17_src/components/opensource/linux/linux-2.6.36/arch/arm/plat-brcm/bcm5301x.c
Oct 03 11:20:04 <zajec> don't look at R7000, because it isn't there
Oct 03 11:21:50 <zajec> Hauke: soc_cca_irq_enable
Oct 03 11:21:59 <Hauke> ok thanks
Oct 03 11:23:55 <zajec> Hauke: AFAIU it uses soc_cca_irq_handler as IRQ 117 handler, then it checks for GPIO IRQ bit and calls soc_cca_gpio_handle if needed
Oct 03 11:24:09 <zajec> soc_cca_gpio_handle does all this polarity magic
Oct 03 11:24:14 <Hauke> yes it looks like the irq domain stuff
Oct 03 11:24:45 <zajec> AFAIR it fakes IRQ using generic_handle_irq
Oct 03 11:25:27 <Hauke> yes
Oct 03 11:26:11 <zajec> for some reason my INTSTATUS register never has SOC_CHIPCOMON_A_INTMASK_GPIO bit
Oct 03 11:26:22 <zajec> i only had SOC_CHIPCOMON_A_INTMASK_UART bit set
Oct 03 11:26:44 <zajec> Hauke: how we should handle BCM4709?
Oct 03 11:26:53 <Hauke> have you seen: /* Enable GPIO interrupts in ChipcommonA */
Oct 03 11:26:53 <zajec> should we create a new DTS file for it?
Oct 03 11:27:09 <zajec> Hauke: i think I did
Oct 03 11:27:14 <zajec> but may need to check it again
Oct 03 11:27:31 <Hauke> how different is bcm4709?
Oct 03 11:28:00 <zajec> well... i don't see any difference
Oct 03 11:28:05 <zajec> except for UART clock aybe
Oct 03 11:29:09 <Hauke> how did the uart clock changed?
Oct 03 11:29:24 <Hauke> bcm4709 has 3 PCIe controllers is that correct?
Oct 03 11:29:48 <zajec> you should ignore that
Oct 03 11:29:56 <zajec> I have BCM4708 with 3 controllers
Oct 03 11:30:02 <zajec> so it's not constant
Oct 03 11:30:38 <Hauke> ok
Oct 03 11:30:40 <zajec> Hauke: https://git.kernel.org/cgit/linux/kernel/git/linville/wireless-next.git/tree/drivers/bcma/driver_gpio.c#n162
Oct 03 11:30:47 <zajec> Hauke: so driver_gpio.c enables GPIO interrupts
Oct 03 11:30:53 <zajec> so it should be working for me
Oct 03 11:30:59 <zajec> still, it didn't
Oct 03 11:31:29 <Hauke> hmm strange
Oct 03 12:25:28 <builder> build #751 of orion is complete: Failure [failed compile_8]  Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/751
Oct 03 12:52:43 <karlp> basic rust on openwrt, whee http://pastebay.net/1501643
Oct 03 12:54:13 <jow_laptop> is it another one of these hipster languages?
Oct 03 12:54:50 <karlp> probably
Oct 03 12:55:25 <karlp> been looking at using something like D/go/? for a while,
Oct 03 12:55:39 <karlp> lua hasn't been as productive as I'd hoped.
Oct 03 12:58:01 <jow_laptop> one major critique about lua has been the lack of a proper ecosystem (like python has)
Oct 03 12:58:11 <jow_laptop> not sure if using even more niche languages will help with that
Oct 03 12:59:45 <builder> build #674 of sibyte is complete: Failure [failed shell_10]  Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/674
Oct 03 13:02:19 <karlp> totally agree
Oct 03 13:08:04 <karlp> file is 1meg anyway, completley useless for running embedded :)
Oct 03 13:43:53 <builder> build #686 of avr32 is complete: Failure [failed compile_5]  Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/686
Oct 03 14:22:29 <zajec> Hauke: can you look at [PATCH][RFC] bcma: fill core details for every device
Oct 03 14:22:50 <zajec> it's needed to fill of_device for ChipCommon
Oct 03 14:23:00 <Hauke> zajec: yes I am currently going through my mails
Oct 03 14:23:02 <zajec> ok :)
Oct 03 14:23:38 <Hauke> Yes I know that we need some changes for chipcommon, I already had a patch prepared, but yours looks better
Oct 03 14:26:24 <zajec> Hauke: do you know if any MIPS BCM47XX patches has been picked?
Oct 03 14:26:40 <Hauke> zajec: no I do not know
Oct 03 14:26:40 <zajec> can't see anything in http://git.linux-mips.org/?p=ralf/upstream-sfr.git;a=summary
Oct 03 14:29:36 <Hauke> I think this tree still misses many other mips patches
Oct 03 14:30:06 <zajec> i'll try to write Ralf
Oct 03 14:30:17 <Hauke> zajec: thanks
Oct 03 14:38:48 <zajec> Hauke: what is BSP?
Oct 03 14:38:50 <zajec> Broadcom BSP
Oct 03 14:39:04 <Hauke> board support packagehttps://en.wikipedia.org/wiki/Board_support_package
Oct 03 16:40:10 <zajec> is there some example of using libuci to iterate over all configs in a file/package?
Oct 03 16:40:29 <jow_laptop> zajec: you mean all sections?
Oct 03 16:40:31 <zajec> I found uci_lookup_next: http://lxr.mein.io/source/uci/uci.h#L321 and I guess I need it
Oct 03 16:40:37 <zajec> yes
Oct 03 16:40:55 <zajec> I want to read all "config interface", one by one
Oct 03 16:41:10 <jow_laptop> http://lxr.mein.io/source/rpcd/uci.c#L508
Oct 03 16:41:41 <zajec> oh... how did i miss that
Oct 03 16:41:43 <zajec> thanks!
Oct 03 18:01:43 <builder> build #777 of at91 is complete: Failure [failed compile_5]  Build details are at http://buildbot.openwrt.org:8010/builders/at91/builds/777
Oct 03 18:04:35 <builder> build #609 of iop32x is complete: Failure [failed shell_10]  Build details are at http://buildbot.openwrt.org:8010/builders/iop32x/builds/609
Oct 03 19:18:11 <zajec> need someone to remind me location of user script
Oct 03 19:18:20 <zajec> i remember there was a place to put user script
Oct 03 19:18:26 <zajec> that will be called on DHCP response
Oct 03 19:18:33 <zajec> when my WAN gets an IP
Oct 03 19:18:39 <builder> build #200 of brcm2708 is complete: Failure [failed compile_4]  Build details are at http://buildbot.openwrt.org:8010/builders/brcm2708/builds/200
Oct 03 19:18:40 <zajec> I can't Google it :|
Oct 03 19:19:30 <zajec> i think dnsmasq or sth got a condition like [ -x /user/dhcp.sh ] && /user/dhcp.sh
Oct 03 19:46:19 <zajec> got it! udhcpc.user!!
Oct 03 22:29:54 <jow_laptop> zajec: it used to be /etc/udhcpc.user
Oct 03 23:21:20 <builder> build #93 of brcm63xx.smp is complete: Success [build successful]  Build details are at http://buildbot.openwrt.org:8010/builders/brcm63xx.smp/builds/93
Oct 03 23:30:47 <builder> build #552 of octeon is complete: Failure [failed compile_5]  Build details are at http://buildbot.openwrt.org:8010/builders/octeon/builds/552
**** ENDING LOGGING AT Sat Oct 04 03:00:00 2014